Aphasia test results. Changes in language performance on the Aachen Aphasia Test (AAT; see panel A) and on the Action CommunicationTest (ACT; see panel B). Thirty individuals with chronic post-stroke aphasia were randomly assigned to one of the two Groups: patients receiving IntensiveLanguage-Action Therapy with 4 hours (Group I) or with 2 hours of daily practice (Group II). All patients went through an initial waiting period (‘baseline’)and two successive training intervals (‘therapy phase’). Each trial phase lasted 2 weeks. Testing took place at four points in Time: 2 weeks before treatmentonset (T0), at treatment onset (T1), after the first training interval (T2) and after the second training interval (T3). Focusing on changes in languageperformance separately for each trial phase [Δ(T1–T0); Δ(T2–T1); Δ(T3–T2)], statistics refer to significant paired-sample t-tests (asterisks embedded in bargraphs) and to a significant Time-by-Group interaction, as revealed by repeated-measures analyses of variance (asterisks displayed above bar graphs;*P